A typical head positioning system for a magnetic disk drive includes a servo loop using a dibit position encoding technique to derive error positioning information. The dibit technique involves the recording of servo information components on two servo tracks located at the sides of a data track, the servo information either being on a dedicated servo surface or embedded into dedicated portions of data surfaces. The two servo information components are prerecorded by two equal-amplitude, sinusoidal waveforms having the same frequency but displaced 90 degrees with respect to each other. The servo information is read by a head which produces a composite dibit signal comprised of servo information components read from both sides of the data track. A clock pulse identifies alternating segments of the composite signal, the first 180 degrees of the composite signal defining a first segment set and the other 180 degrees defining a second segment set. The peak-to-peak voltage of the first segment set is compared to the peak-to-peak voltage of the second segment set. If the head is centrally positioned with respect to a desired data track, the peak-to-peak voltage of the two segment sets will be equal. As the head moves off position with respect to the desired data track, the amplitude of one segment set will change with respect to the amplitude of the other segment set. This amplitude imbalance produces an error positioning signal which is fed back to the head positioner to reposition the head until the amplitudes again become equal. The direction of repositioning is determined by whether the head is being positioned over an odd or even numbered data track.
For proper operation of one type of servo preamplifier and demodulator chain, the dibit signal appearing at the output of the head must have its polarity determined according to the data track being sought as above-explained, be amplified by an amplifier having a substantially constant output voltage regardless of input level variations, and be separated into first and second segments in accordance with alternating states of an external clock signal. Conventional servo systems accomplish the above with various types of circuitry well-known in the art. A first-stage preamplifier having adjustable gain is usually provided by an integrated circuit having gain controlled by some version of voltage or current controlled resistor, such as a field effect transistor (FET). This type of resistor is very nonlinear for higher input signals and the amplitude of an input signal has to be small to reduce adverse effects of this nonlinearity. Thus, automatic gain control is frequently implemented in the first stage even though at this point it might reduce the signal-to-noise ratio and/or cause distortion. The polarity switch is conventionally implemented by various techniques using transformers and/or MOS switches. These circuits are costly and introduce errors because of limited bandwidth, stray fields and non-ideal characteristics of the switch and/or transformer. In order to eliminate some of the above problems, one approach has been to implement two separate channels, one for each polarity. Such a solution at best is more expensive. The signal separator is usually constructed utilizing a combination of diode switches activated by logic circuits and a clock pulse. In order to implement the above-described functions, conventional servo preamplifier and demodulator circuits utilize many different types of components which are expensive both in terms of procurement cost and fabrication time.